REF-1405 · Watson et al., 2018 · Journal of Bone and Mineral Research · Paper · PubMed

Can heavy lifting build bone after menopause, even with low bone mass?

In 101 women (~65 y, low bone mass), 8 months of supervised 5×5 lifting above 85% 1RM plus impact raised spine density 2.9% vs a 1.2% loss; one minor back spasm.

Main caveat: Read from the abstract only. Registration, primary outcome, randomisation, allocation concealment, assessor blinding, dropouts, the analysis population, the…
fairly strong evidenceMeasured on non-athletes and recreational people

Watson and colleagues start from the principle that bone responds best to high-magnitude strains applied at high rates, and note that the training which applies them, high-intensity resistance and impact training (HiRIT), is not traditionally recommended for people with osteoporosis because of a perceived high fracture risk.

The LIFTMOR trial set out to test its efficacy on markers of fracture risk, and to monitor adverse events, in postmenopausal women with low bone mass. 101 women (65 ± 5 years, 161.8 ± 5.9 cm, 63.1 ± 10.4 kg) with a T-score below −1.0, screened for conditions and medications that influence bone and physical function, were randomised to 8 months of twice-weekly, 30-minute supervised HiRIT (resistance work of 5 sets of 5 repetitions at more than 85% of one-repetition maximum, with impact training; the abstract does not describe the impact exercises) or to a home-based, low-intensity exercise programme (CON).

HiRIT (n = 49) beat CON (n = 52) for lumbar-spine density (+2.9 ± 2.8% vs −1.2 ± 2.8%), femoral-neck density (+0.3 ± 2.6% vs −1.9 ± 2.6%), femoral-neck cortical thickness (+13.6 ± 16.6% vs +6.3 ± 16.6%), height (+0.2 ± 0.5 vs −0.2 ± 0.5 cm) and all functional measures (timed up-and-go, functional reach, five times sit-to-stand, back and leg strength; p < 0.001).

Compliance was 92 ± 11% with HiRIT and 85 ± 24% with CON; one adverse event was reported (a minor lower-back spasm in a HiRIT participant, who missed 2 of 70 sessions). The authors conclude that their brief programme enhanced indices of bone strength and function and, "contrary to current opinion", was efficacious and safe under highly supervised conditions.

Our reading: this is the base's first trial of a concrete bone-loading prescription in postmenopausal women, and the first to show that heavy loading was tolerated in women with low bone mass; it compares a heavy-plus-impact programme with a low-intensity home programme, so it shows the package beat light home exercise, not which part (the heavy lifting, the impact or the supervision) did the work.

A meta-analysis already in the base found no significant density difference between heavier (70% 1RM and above) and lighter resistance training in adults over 50 (REF-1086, A), which is a different comparison (two resistance intensities, no impact component).

Protocol

Design: randomised controlled trial, two parallel arms, 8 months; pre- and post-intervention testing. Registration, randomisation method, blinding of assessors and the analysis population are not in the abstract.
Who: n = 101 postmenopausal women with low bone mass (T-score < −1.0), aged 65 ± 5 years, height 161.8 ± 5.9 cm, body mass 63.1 ± 10.4 kg; screened for conditions and medications that influence bone and physical function. Time since menopause, current or past hormone therapy or osteoporosis medication, and training history are not in the abstract.
HiRIT (n = 49): supervised, twice a week, 30 minutes, 8 months; resistance exercise at 5 sets of 5 repetitions at more than 85% of 1RM, combined with impact training (exercise selection and impact loads not in the abstract).
CON (n = 52): a home-based, low-intensity exercise programme (content not in the abstract).
Outcomes: lumbar-spine and proximal-femur BMD (DXA); femoral-neck cortical thickness; height; functional performance: timed up-and-go, functional reach, five times sit-to-stand, back extensor and leg strength. Adverse events and compliance were monitored.
Funding and interests: not available from the abstract; PubMed classes the trial as "Research Support, Non-U.S. Gov't" without naming the funder, and Crossref lists none. Affiliations as printed: Griffith University and Menzies Health Institute Queensland (five authors), and The Bone Clinic, Brisbane (Weis; Beck also lists it). No conflict-of-interest statement is available from the abstract or the PubMed record; no device maker, federation or data platform is named.

Key Results

Outcome (8-month change)HiRIT (n = 49)CON (n = 52)Between groups
Lumbar-spine BMD+2.9 ± 2.8%−1.2 ± 2.8%p < 0.001
Femoral-neck BMD+0.3 ± 2.6%−1.9 ± 2.6%p = 0.004
Femoral-neck cortical thickness+13.6 ± 16.6%+6.3 ± 16.6%p = 0.014
Height+0.2 ± 0.5 cm−0.2 ± 0.5 cmp = 0.004
Timed up-and-go, functional reach, 5 × sit-to-stand, back and leg strengthimproved more—all p < 0.001
Compliance92 ± 11%85 ± 24%—
Adverse events1 (minor lower-back spasm; 2 of 70 sessions missed)none reported—
The ± figures: the spread is identical in the two arms for every outcome (2.8, 2.6, 16.6, 0.5), which suggests model-derived or pooled values rather than raw SDs (our reading; the abstract does not say what they are). No confidence intervals are given.
Total hip: proximal-femur density was measured; the abstract reports only the femoral neck.
Not in the abstract: dropouts and the number analysed, intention-to-treat or not, effect sizes for each functional test, and the starting loads.

Mechanisms

Measured: density rose at the lumbar spine and held at the femoral neck with HiRIT while it fell with the home programme; femoral-neck cortical thickness rose more; height was kept.
Authors' interpretation: optimal osteogenic loading needs high-magnitude strains applied at high rates, which HiRIT supplies and low-intensity exercise does not; the fear of fracture behind the usual advice against such loading in osteoporosis was not borne out under supervision.
Inference: this matches the base's bone principle — bone answers the size and novelty of the load, a few loads at a time (PHYS-37); with impact and heavy lifting in one arm and neither in the other, the trial cannot apportion the gain between them.

Practical Application

Can heavy lifting build bone after menopause, even with low bone mass? In 101 women (~65 y, low bone mass), 8 months of supervised 5×5 lifting above 85% 1RM plus impact raised spine density 2.9% vs a 1.2% loss; one minor back spasm.
The tested dose: twice a week, 30 minutes, supervised, 5 sets of 5 at more than 85% of 1RM with impact training, for 8 months (A-, RCT, from the abstract). Not from this paper; the base's own practice: add impact or heavy lifting outside the sport for bone (PHYS-37); this trial is the concrete shape such sessions can take for a masters woman, started under a coach and built up to these loads (inference: tested only in non-athletes with low bone mass).
Safety was shown under close supervision only: one minor back spasm in 49 women training with heavy loads; the authors stress "highly supervised conditions" and screened, otherwise healthy women. Whether self-coached heavy lifting is as safe for a woman with osteoporosis was not tested.Not from this paper; the base's own practice: a woman who knows she has osteoporosis or has had a fragility fracture starts such a programme with her doctor's agreement and a qualified coach.
Creatine is not the bone tool (conflict with [REF-530]): the base's narrative entry REF-530 (B+) reports that postmenopausal high-dose creatine with resistance training has "favourable effects on bone", which SYNTH-67 repeated; since 2026-10-28 both set that claim against the evidence below, and the base does not present creatine as a bone treatment (SYNTH-96).One RCT and one meta-analysis of five RCTs, not yet in the base, point the other way: a 2-year RCT in 237 postmenopausal women (Chilibeck 2023) and a meta-analysis of 5 RCTs in adults over 50 or postmenopausal (Forbes 2018) found no effect of creatine added to resistance training on bone density, although the 2-year RCT reported better femoral-neck geometry (section modulus, buckling ratio) (from their abstracts, PMID 37144634 and 29740583; not read in full).The bone stimulus that has been tested and worked in this batch is loading, as here.
Heavy vs lighter lifting: a meta-analysis in the base found no significant density difference between resistance training at 70% 1RM and above and lighter training in adults over 50 (REF-1086, A); LIFTMOR compared a heavy-plus-impact programme with light home exercise, so the two do not contradict each other, and neither isolates the impact component.
What it is good for as a source: a randomised, concrete bone-loading prescription in postmenopausal women and evidence that, supervised, it was tolerated with low bone mass. What it does not show: fracture reduction; effects in women with normal density, in perimenopause or in athletes; which component worked; durability after the 8 months.

Limitations

Read from the abstract only. Registration, primary outcome, randomisation, allocation concealment, assessor blinding, dropouts, the analysis population, the impact exercises and the control programme could not be checked.
Surrogate outcomes. Areal BMD, femoral-neck geometry, height and function — no fractures; 8 months, no follow-up after the programme.
Comparator. A low-intensity home programme, not a matched-attention or matched-volume resistance programme; supervision, impact and load differ together between arms.
Who. Postmenopausal women aged about 65 with low bone mass, screened for conditions and medicines affecting bone; not athletes, not perimenopausal, not women with normal density.
Reporting. The abstract reports "only one adverse event" in its results and, in its conclusion, that HiRIT "induced no adverse events"; the identical ± values across arms are not explained; no confidence intervals.
Funding and interests: not available from the abstract; "Research Support, Non-U.S. Gov't" (PubMed, funder unnamed). Two authors list The Bone Clinic, Brisbane, as an affiliation (as printed in the PubMed record); no conflict-of-interest statement is available to say whether that is a commercial interest in the programme.
How this entry was checked and changed
Why this level

(from the abstract) — a parallel RCT well above the rubric's 40, and above 100 in total, but held below A because only the abstract was read (closed access: the publisher's free PDF returned a Cloudflare check and the OpenAlex-listed Griffith repository record exposes no open file), so trial registration, the primary outcome, randomisation and blinding details, the analysis population and dropouts cannot be confirmed; the outcomes are density, geometry and function, not fractures